Monitoring to large electrical water pumps.

[mramos1]

Our building has two 480V 3 phase motors to pump water into the building.

They both have a manual knob (turned 45 degrees) to force them on as the automatic controller is dead (looks like no 5VDC on the board). So we have ran many year on one pump and keep the other for a backup.

Well, we had a breaker/contactor trip and lost pressure a couple weeks ago and the chilling tower on the roof was not getting water but we caught it in time.

At $10,000US for a new panel, we came up with the idea to monitor the pumps and message when there is a problem.

I wanted to share my ideas and get some input.

Vibration is out and as well as noise levels. However, a coil around the power lines to monitor the power came to mind (like the old tachs for cars).
I would not have to hack into the system that way.

I would have a PIC monitor the current on both pumps. If I see not power for 20 minutes or so, sound an alarm.

I have thought about an old PC for the RTC and TCP/IP stack. I would have less false alarms at night when the building is closed and one the weekends as well I could SMS a couple phones.

Or maybe use an X-bee pro to send a signal to my computer in my office to SMS me. I have had really good luck with the maxstream parts and distance.

There is a presure sensor going into the controller and probably works, but it did not look easy to access. But that would probably be a better way to do it?


Any better ideas (besides spending $10K)

 

[Pommie]

A coil around the power lines won't work unless you can separate the individual phases. If you go around all phases then the net magnetic field is zero. Can you not just measure the actual voltage.

A better idea might be to fix the existing board.

Mike.

 

[mramos1]

I am waiting on a company to call me back to tell me if while they are in hand/manual mode if the controller is out of the circuit. If so, then I will attempt to see why there is no 5VDC on the board.

But $10,000 for something that simple, just can not see spending the money.

Both motors have an electrical box I can grab a phase from.

Maybe I should re-visit vibration at least for the "warning they are not running" part.

 

[Leftyretro]

Our building has two 480V 3 phase motors to pump water into the building.

They both have a manual knob (turned 45 degrees) to force them on as the automatic controller is dead (looks like no 5VDC on the board). So we have ran many year on one pump and keep the other for a backup.

Well, we had a breaker/contactor trip and lost pressure a couple weeks ago and the chilling tower on the roof was not getting water but we caught it in time.

At $10,000US for a new panel, we came up with the idea to monitor the pumps and message when there is a problem.

I wanted to share my ideas and get some input.

Vibration is out and as well as noise levels. However, a coil around the power lines to monitor the power came to mind (like the old tachs for cars).
I would not have to hack into the system that way.

I would have a PIC monitor the current on both pumps. If I see not power for 20 minutes or so, sound an alarm.

I have thought about an old PC for the RTC and TCP/IP stack. I would have less false alarms at night when the building is closed and one the weekends as well I could SMS a couple phones.

Or maybe use an X-bee pro to send a signal to my computer in my office to SMS me. I have had really good luck with the maxstream parts and distance.

There is a presure sensor going into the controller and probably works, but it did not look easy to access. But that would probably be a better way to do it?


Any better ideas (besides spending $10K)

This is a common application in the process control industry where you have a pair of pumps. Typically one is running continuously and the 2nd is on "automatic standby" ready to start if the running pump fails.

Monitoring the current draw is a way to activate a standby pump but typically that's not how it's done. For instance what if the impeller wheel failed in the pump or the motor/pump coupling failed? the motor would still be running but no flow would be generated. It's best to control via the primary process variable, the water flow or pressure.

There is usually a flow sensor or pressure sensor switch installed in the piping that is common to the discharge line of both pumps. There is usually a check valve (one way flow valve) on the discharge of both pumps so that flow is not directed backwards into the standby pump, but after the check valves the flow (and pressure) is common to both pumps. So if the pressure or flow sensor switches then, that is wired to the start circuit of the standby pump.

There is usually a field switch for each motor with three positions, on, off and automatic. That way either pump can be the "lead" pump by putting it in on and the other in auto. That allows pump run time to be balanced by switching their roles every week or month, and also allows one to operate both pumps on or off for testing or maintenance.

There is usually no need for a complex controller for this type of application, usually just a couple of relays if you want low voltage control to the field switches and sensor switch but sometimes they just use a 120vac control power from the motor starters to the field on/off/auto switches and sensor.

Lefty

 

[Joe McGivern]

I will, if I may, ask you some questions. This is simply to try to understand some of the basic details of your installation.

How large are "large electrical water pumps"in terms of Kw?


Are the drive motors, (1)Direct On Line (2) Star/Delta or (3) Variable Speed Drive controlled?

Do they pump to a header tank/reservoir or is it a pressurised system?

How is the water supplied to the pumps. Is it sucked from a tank at a level below the pumps or do they have a "positive feed"?

JMcG